Skip to content

Benchmark

Arena ELO - English

People compare model answers and choose which they prefer. Higher scores are better.

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.26

Score distribution

1,327.7
Mean
1,373.1
Median
138.5
Std dev
1,119.5
10th percentile
1,468.7
90th percentile
845.9
Min
845.91,532
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.26 for chat models, measured over 366 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
251Yi Lightning· LMArena data1,329.8
252Hunyuan Turbo0110· LMArena data1,328.8
253Claude Sonnet 3.7Thinking 32K, 20250219· LMArena data1,326.4
254Gemini 1.5 Pro002· LMArena data1,322.6
255Qwen2.5-Plus1127· LMArena data1,322
256GPT-5 nanoHigh· LMArena data1,321.4
257Gemma 3n E4B IT· LMArena data1,321.4
258Grok 22024-08-13· LMArena data1,320.7
259Granite 4.2 3B· LMArena data1,319.5
260Granite 4.1 8B· LMArena data1,315.8
261Llama 3.1 Nemotron 70B Instruct· LMArena data1,314.9
262GPT 4o2024-05-13· LMArena data1,312.4
263Llama 3.1 405B InstructBF16· LMArena data1,311.8
264Claude Sonnet 3.720250219· LMArena data1,310.1
265DeepSeek V2.51210· LMArena data1,309.6
266Llama 3.1 405B InstructFP8· LMArena data1,307.9
267Claude 3.5 Sonnet20241022· LMArena data1,307.1
268Athene V2 Chat· LMArena data1,306.9
269Hunyuan Large2025-02-10· LMArena data1,305.1
270Llama 4 Scout· LMArena data1,304.3
271Llama 3.3 70B Instruct· LMArena data1,303.5
272GPT OSS 20b· LMArena data1,303.3
273GPT-4.1 nano2025-04-14· LMArena data1,303.2
274Llama 4 Maverick· LMArena data1,302.4
275GPT-4o mini2024-07-18· LMArena data1,300.6
276Mercury· LMArena data1,300.5
277Gemma 3 4B IT· LMArena data1,300.4
278Molmo 2 8B· LMArena data1,299.2
279Grok 2 mini2024-08-13· LMArena data1,298
280GPT 4o2024-08-06· LMArena data1,297.4
281GLM 4 Plus· LMArena data1,296.7
282Mistral Small 3.1 24B Instruct2503· LMArena data1,294.4
283Llama 3.1 70B Instruct· LMArena data1,293.7
284Qwen Max0919· LMArena data1,292.6
285GPT-4 Turbo2024-04-09· LMArena data1,291.2
286Mistral Large 22407· LMArena data1,289
287Gemini 1.5 Flash002· LMArena data1,288.6
288Athene 70B0725· LMArena data1,288.1
289Claude 3.5 Sonnet20240620· LMArena data1,287.3
290DeepSeek V2.5· LMArena data1,284.3
291Mistral Large2411· LMArena data1,283.9
292Qwen2.5 72B Instruct· LMArena data1,282.6
293Gemini Advanced0514· LMArena data1,281.6
294Hunyuan Standard2025-02-10· LMArena data1,281.5
295Hunyuan Large Vision· LMArena data1,281.1
296Gemini 1.5 Pro001· LMArena data1,279.3
297Llama 3 70B Instruct· LMArena data1,277.4
298GPT-4 Turbo Preview1106· LMArena data1,277.1
299GPT-4 Turbo Preview0125· LMArena data1,276.1
300Nova Pro 1.0· LMArena data1,276.1

Showing 251–300 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)